Alexandria and the Moon: Addenda et Corrigenda more

ALEXANDRIA AND THE MOON ADDENDA ET CORRIGENDA Second edition: 19 March 2012 This note covers additions and corrections to my study of the Egyptian Macedonian calendar published as Studia Hellenistica 52. It will be updated from time to time as material comes to my attention. Addenda 1. Willy CLARYSSE (pers. comm., 4 August 2011) has drawn my attention to the contract register CPR XVIII. The editor of CPR XVIII, Bärbel KRAMER, assigned these contracts to Ptolemy III but other scholars such as Klaus MARESCH date them to Ptolemy IV. The register is dated to Pharmouthi year 16 = FY 17. The surviving dates of the contracts listed in the register are Dystros (2), Artemisios (7), Daisios (1), Panemos (3), Hyperberetaios (9) and Dios (1), all in year 16. If the king is Ptolemy III, then Pharmouthi year 16/FY 17 = 18 May – 16 June 231 ≈ 17.II 231/0. As reconstructed here, the latest contract date of Dios year 16 ≈ 16.XII2 232/1 = c. 18 March – 15 April 231 is two months before the register date. (R. ZIEGLER, ZPE 125 (1999) 211f., assigns the Macedonian months to the previous lunations, but this cannot be correct if the analysis of the double dates of pPetrie III 53(s) and SB XVIII 13255 presented in section V.3 is correct.) If the king is Ptolemy IV, then Pharmouthi year 16/FY 17 = 12 May – 10 June 206 ≈ 4.II 206/5 = Hyperberetaios year 16 in Alexandria, as reconstructed here. Additionally, on the assimilated calendar Pharmouthi = Hyperberetaios at this time. On this basis, the contract CPR XVIII.6, dated Dios year 16, postdates the register. The possibility that a contract written at the very start of Dios was included in the registered list at the last minute cannot be excluded. However, the calendrical evidence favours assigning the register to Ptolemy III rather than Ptolemy IV. None of the dates in this register provide any additional fixed points or cross-checks to the reconstruction of Table 12. 2. p. 34 n. 12: On techniques for converting Babylonian dates to Egyptian ones, see now STEELE (2011). The technique described for computing exact Egyptian equivalents of Babylonian dates of lunar eclipses using the Saros cycle, given one known conversion, allows the equivalent dates of any intermediate Babylonian month to be calculated rapidly from a lunar table without processing the entire table. 3. p. 51 n. 60: On certain characters on the tetradrachm Svoronos 1486, interpreted by Richard HAZZARD to represent year 33 (Egyptian) of Ptolemy VI, see now Catharine LORBER apud BLASIUS (2011) at 176. She suggests that the characters should be read ΙΔΓΑ (where each character is reversed), not LΛΓΑ, and that they represent the marks of two mint masters, ΔΙ and ΑΓ. She notes that one of these is independently 19 MAR 2012 1 ALEXANDRIA AND THE MOON documented (as ΔΙΟ) in 144/3-142/1, shortly after the Seventh Syrian War, and that a similar dual marking is seen a few years earlier in coinage of Alexander Balas. 4. p. 70 n. 8: On the double date partially known through the very fragmentary Greek text of iGLouvre 2: The content of this synodal decree is now known in full, through the discovery (EL-MASRY (2010)) and publication (El-MASRY ET AL. (2012)) of a stele giving the hieroglyphic and demotic versions. Prof. THISSEN kindly informs me (pers. comm., 11 March 2012) that its date is now confirmed as Gorpiaios 13 = Phaophi 12 year 5 of Ptolemy III = 243/2. Since this is a day-exact double date, it may be added to the corpus as follows: a. p. 232 Table 5: after the row for PCZ III 59327 insert: Year 5 Document [TBD] Double Date Gorpiaios 13 = Phaophi 12 Day 1 (Egyptian) Thoth 30 Day 1 (Julian) 21 Nov. 243 Last visibility 20 Nov. 243 Delta -1 First visibility 23 Nov. 243 Delta +2.5 Lunation 5.IX 243/2 This Gorpiaios nominally began on the Egyptian lunar day psDntyw, the first day of the Egyptian lunar month. However, the double dates in the Canopic and Rosetta decrees do not show the same lunar alignment, so this is probably coincidental. b. p. 238 Table 11: replace the row for pTebt. III.1 814 by: Source [TBD] pTebt. III.1 814 Macedonian month / year Gorpiaios 5 Pt. III Gorpiaios 8 Pt. III Julian day 1 [Nominal] 21 Nov. 243 20 Nov. 240? Lunation 5.IX 243/2 8.IX? 240/39 No of intercalations 3 1? Lunisolar delay +1 0? Notes Synodal decree Or one month earlier c. p. 242 Table 12: replace the row for lunar year 5 by: 4/5 5 243/2 Aud. Per. Dys. Xan. Art. Dai. Dai. Pan. Loi. Gor. Hyp. Dios Ap. Aud. [TBD] Corrigenda 1. p. 206 lines 7-8: Farhad ASSAR (pers. comm. 15 October 2011) notes that my estimate of the minimum time from Xandikos 3/23 to Panemos 10 SEM 119 as 45 days is incorrect: it should be 74 days, since there were two regular months between Xandikos and Panemos, not one. As Dr ASSAR notes, this apparently minor error effectively reverses the model of Seleucid intercalation developed in Appendix C.2: i) Since 74 days greatly exceeds the estimated time for delivering the message of SEG XIII 592 from Antiochus III to Menedemos by horse, this model for delivering the decree, which is the basis of the analysis of Appendix C.2, cannot be correct. The three routes considered give the following speeds for a transmission delay of 74 days: i. Ephesus – Ecbatana (1600 miles): 2 21.6 miles / day 19 MAR 2012 ADDENDA ET CORRIGENDA (Second edition) ii. Pisidia – Ecbatana (1300 miles): 17.6 miles / day iii. Pisidia – Sardis – Ecbatana (1700 miles): 23 miles / day If we assume an intercalary month, i.e. c. 104 days, then the transmission speeds become 15.4, 12.5 and 16.3 miles / day respectively. ii) We may reasonably assume that the decree of SEG XXXVII 1010, issued probably in Ecbatana on Dystros 23 and republished in Sardis on Artemisios 3 of SEM 103, was transmitted in the same way. At the maximum rate of 23 miles a day, the decree traveled at most 920 miles if there was no intercalation, and about 1600 miles if there was one. Since the distance from Ecbatana to Sardis is about 1500 miles, SEM 103 must have been intercalary, with either a Dystros or a Xandikos embolimos. iii) SEM 103 = 210/9 is a cycle year 19, which is ordinary on both the autumn cycle developed in Appendix C and the alternative suggested on p. 209 n. 35. Therefore Seleucid intercalation did not follow the autumn cycle developed in Appendix C. If Seleucid intercalation used a 19 year cycle, it was most probably a spring cycle. iv) Knowing SEM 103 was intercalary allows us to refine our estimate of the transmission speed: 1500 miles in about 68 days gives a speed of about 22 miles a day, consistent with the available data for military convoys on foot. Hence there was no intercalary month between Xandikos and Panemos SEM 119. v) SEM 119 = 194/3 is a cycle year 16, which is intercalary on a spring cycle. On the Parthian Macedonian calendar, the intercalary month would be Dystros embolimos, which is before the date of the letter in SEG XIII 592. vi) We may conclude that there are reasonable grounds for supposing that the Seleucid Macedonian calendar of Antiochus III was the same as the later Parthian Macedonian calendar, in both its intercalary cycle and its solar alignment. This result affects my conclusions on a number of issues: i) Is the Seleucid solar alignment of Antiochus III related to the Ptolemaic solar alignment of the first two decades of Ptolemy II? In section VI.7 I argued that the Ptolemaic alignment was unrelated to the later Parthian solar alignment since all evidence considered, including this Seleucid data, indicated that the Macedonian calendar outside Egypt and Parthia held to the alignment of Alexander’s time. However, if the Parthian alignment was already in use by Antiochus III the possibility of a common origin must be reconsidered. ii) When did the change of solar alignment happen? There are four possibilities: a. There was no change: i.e. the alignment of Antiochus III, like that of first half of the reign of Ptolemy II, reflected and inherited the normal solar alignment of the Argaead calendar. 19 MAR 2012 3 ALEXANDRIA AND THE MOON If this theory is correct, the two clear Argaead synchronisms we possess, for Alexander’s birth and death, together with the strong likelihood that Philip’s death in late summer 336 fell in early Dios, would all have to result from phase variance in which a Macedonian intercalation (e.g. after Hyperberetaios) lagged behind a Babylonian intercalation, with the synchronism occurring between the two. Alexander’s birth and death both fell shortly after an Addaru II, so could be consistent with this model. On the other hand, Philip’s death was in the second of two ordinary Babylonian years. Additionally, the death of Cassander in Artemisios embolimos 298/7 falls in an ordinary year on the Parthian calendar. While the solar alignment of the lunation corresponding to this month is not known, this intercalary discrepancy shows that the Parthian intercalary cycle does not reflect that of fourth century Macedon, and it is therefore unsafe to assume a common solar alignment without positive evidence. b. Between the death of Alexander and Seleucus’ calendar reform. The shifted alignment seen in the first half of the reign of Ptolemy II may reflect an earlier shift which was inherited by both the Ptolemies and the Seleucids, but was not followed in Macedon. It is unclear when such a shift could have occurred. c. As part of Seleucus’ calendar reform. This is the most obvious solution, but there is no positive evidence for it. As discussed next, the “Chaldean” dates of the Almagest are arguably prima facie evidence against it. A further circumstantial argument against it is the apparent failure to maintain this alignment in post-Seleucid territories, particularly in Anatolia. d. Between Seleucus’ calendar reform and the reign of Antiochus III. The “Chaldean” dates of the Almagest are prima facie evidence in favour of this, and might even suggest that the shifted alignment was introduced by Antiochus III himself. However, even though it seems to me most likely that the Macedonian form of these dates is Graeco-Babylonian in origin, as discussed in section III.1, this cannot be proven. Nor is it certain that the Macedonian dates are contemporary with the observations. If instead they represent a conversion or translation made in a territory using the more familiar solar alignment at a later time, they have no evidentiary value for Seleucid calendar history. iii) Where did the autumn cycle come from and when? The epigraphic and numismatic calendrical evidence considered in Appendix C from Macedon, Pontus, and post-Seleucid Anatolia, along with the (more debatable) evidence of Malalas for post-Seleucid Antioch, remains consistent both with an autumn cycle and the solar alignment of Alexander’s time. The death of Cassander in Artemisios embolimos 298/7 and the assignment of SEG XXVII 279 to Antigonus II, if correct, are consistent with use of an autumn cycle in Macedon as early as the start of the third century, though they 4 19 MAR 2012 ADDENDA ET CORRIGENDA (Second edition) do not attest to its solar alignment. Aside from the “Chaldean” dates, the earliest evidence supporting the solar alignment of this cycle is the first century date of Xandikos embolimos of SEG XXXI 1046. However the first direct epigraphic attestation is given by OGIS 458, from 8 B.C. In addition to modifying the discussions in section VI.7 and Appendix C, the following corrections apply to the tables: i) Table 16: Delete rows for SEG XXXVII 1010 and OGIS 224 + SEG XIII 592 ii) Table 17: i. Retitle as “Seleucid, Parthian and Gandharan Intercalary Data” ii. Insert the first two rows as follows: Source SEG XXXVII 1010 OGIS 224 + SEG XIII 592 Location Mysia Phrygia?+Media Type Intercalary year Intercalary year? Date SEM 103 SEM 119 Julian Cycle 210/09 194/3 19 16 Spring Y Y Autumn N N? Notes Dystros II or Xandikos II No Xandikos II Errata • p. XXI entry for DEPUYDT (forthcoming) line 3: “ET AL,” => “ET AL.” • p. 114 3rd para line 1: “the Kallixeinos” => “that Kallixeinos” • p. 116 2nd para line 1: “binennial” => “biennial” • p. 123 line 3: “.118” => “118.” • p. 179 n. 2 line 2: “KOENEN (1988) II 209” => “KOENEN (1988)” • p. 188 2nd para line 10: “30 B.C.” => “31 B.C.” • p. 195 line 4: “two” => “three” • p. 200 n. 8 line 1: “Era,” => “Era, and” • p. 236 Table 11 last row: “+6” => “+5” Additional Bibliography A. BLASIUS, “Antiochos IV. in Ägypten – Ptolemaios VI in Syrien: Die späte Rache des Pharao?!”, in A. JÖRDENS & J. F. QUACK (eds), Ägypten zwischen innerem Zwist und äußerem Druck: Die Zeit Ptolemaios’ VI. bis VIII. Internationales Symposion Heidelberg 16.–19.9. 2007 (Wiesbaden, 2011), 161-185 Y. EL-MASRY, “Recent Excavations at El-Khazindariya in the 9th Nome of Upper Egypt”, in A. NUR EL-DIN & B. S. EL-SHARKAWY (eds.), The Horizon: Studies in Egyptology in Honour of M. A. Nur El-Din (10-12 April 2007) (Cairo, 2010), 179-196 Y. EL-MASRY, H. ALTENMÜLLER & H.-J. THISSEN (eds.), Das Synodaldekret von Alexandria aus dem Jahre 243 v. Chr. (Hamburg, 2012) J. M. STEELE, “Visual Aspects of the Transmission of Babylonian Astronomy and its Reception into Greek Astronomy”, Ann. Sci. 68 (2011), 453-465 19 MAR 2012 5